3C24 new Build
- Thread starter Wntrmute2
- Start date
I must be the only person on the planet who doesn't run AC filament wiring along the floor of the chassis. I keep components down low, and I run filament wires overhead. Everywhere there's a tube socket, I make a drop. Think about a factory with AC outlets hanging down from the ceiling. This approach does two good things. It keeps most of the filament run away from components, and it prevents the imbalance that occurs when a balanced pair (twisted or parallel) is laid against a metallic structure. I've used this approach successfuly in nearly everything I've built, including one very high gain guitar amp with cascaded 12AX7s.
Jack
Jack
That would be the virtual center cap. You can use 220-1K resistors. I have them drawn going to the audio ground which is fine, and in this application the chassis should be OK as well.
The best example of this is the guitar amp I built. It uses twisted filament wiring, so it's easy to distinguish it from everything else. I don't have pictures of that on this computer though, and I'm not sure that I took a good shot of the completed underside. I'll look on my other computer tomorrow.
I have one example on this machine, it's the recently completed 30W/ch pentode amp. However, I used parallel wires for most of the filament string, so it's harder to see. Also, the chassis was less than 2" tall, so the elevation of the wiring isn't so obvious. I'll post it, but I don;'t know if it will really serve all that well to illustrate the principle I'm talking about.
I found a pic of the other amp...
Cool. I think I'll put them closer to the PT as the area by the last tube in the heater string is crowded.
I'm going to redo the heater wiring as I'm not really happy with it running parallel with the other wiring and components.
Hum and noise pickup can be a funny thing. Take a look at the series resistor-diode at the left side of the turret board in the second pic that I posted. This is the bias rectifier for the output tubes. These two components were originally in reverse order. The cathode side of the diode (connected to a power transformer tap) was closest to the front panel and was inducing 60 Hz noise into the nearby pots. The end of the resistor that's near the panel now is bypassed, so it has no AC. This simple change completely eliminated the noise.
Jack
Jack
The work you're doing looks fine to me. I don't think you'll have any issues with the layout. Just remember that the ground busses should be connected back to a common point in the PS (star ground).
What about the notion that the audio bus should ground to the chassis at the input jack? The audio bus is one continuous piece of wire from the PS caps to the "network" I posted.
I never ground the input jacks to the chassis. I either mill a G10 insulating panel for this purpose or use nylon shoulder washers to fit Switchcraft phono jacks. My current 211 project will probably use Amphenol insulated jacks, the type that are mounted in a square receptacle. IMO, grounding an input jack to the chassis is just asking for a ground loop. I only ground the other end of the cable, and I try to position that connection as close to the input tube grid as possible. The purpose of a star ground is to prevent high current devices like output tube cathodes from drawing current through the same conductor(s) being used for small signal ground current. I'm not sure whether your ground bus accomplishes that.
I don't think my bus will accomplish that. The filament of the 3C24 will ground to the bus via the ma meter
The only thing I can tell you is that I only use a ground bus like that when it's convenient for a number of higher-level components. For example, I might use the same bus to ground output stage screen bypass caps, cathode resistors, high-level B+ decouplings, etc. But all the low level stages like input and driver tubes (and the final B+ decoupling caps at those stages) each have their own local ground system connected to the star at the power supply with a single wire. This approach - plus not using the chassis for ground anywhere except a single point at he star) means I never have to go back and search the grounds for hum or ground loops.
I have used the ground bus approach and the star approach, and I tend to at least build power supplies with a bus based approach so ripple current stays in a very confined location.
I do tend to make the audio ground to chassis connection fairly close to the first stage.
I do tend to make the audio ground to chassis connection fairly close to the first stage.
IME that's risky. The third wire ground from the mains usually carries a lot of noise and HF junk. The farther away the connection is made from the last PS cap, the greater the possibility it will modulate the low-level stages. Maybe this doesn't matter in a practical sense, as long as the path from that point to the PS is very low impedance (short heavy gauge wire). I just like to avoid this possibility.
You can always use a ground breaker rather than a piece of solid wire to give some separation if desired. That has been quite effective in linestages. For a DHT power amp, a lot of these choices have consequences that are well buried by the amp's noise floor.
Similar threads
